The goal of this proposal is to further evaluate the molecular origins of in vitro growth characteristics in juvenile chronic myelogenous leukemia (JCML) with particular emphasis on the relation to therapeutic and in vitro responses to 13-cis retinoic acid (CRA). This malignancy of early infancy and childhood is characterized by prominent leukocytosis with monocytosis, anemia, thrombocytopenia, hepatosplenomegaly, elevated fetal hemoglobin, absence of the Philadelphia chromosome, and usually a normal karyotype. JCML is a very different disease than the adult form of CML and carries a very poor prognosis. We have described previously several in vitro growth characteristics of JCML granulocyte-macrophage progenitors (CFU-GM) including: 1) "spontaneous" growth of peripheral blood CFU-GM at low cell densities in the absence of exogenous growth factors, 2) almost complete abrogation of "spontaneous" growth by prior adherent cell (monocyte) depletion or co-culture with CRA, 3) Granulocyte/macrophage colony stimulating factor (GM-CSF) was identified as the principal regulator of in vitro myeloproliferation in JCML, 4) GM- CSF production by JCML monocytes was not consistently excessive, thus paracrine mechanisms seemed unlikely, and finally, 5) selective GM-CSF hypersensitivity of JCML CFU-GM was demonstrated and appears to be the pathophysiologic mechanism behind the myeloproliferation in JCML. Further, we have reported, through a pilot study, that JCML patients respond clinically to CRA. This clinical response is accompanied by a reduction in the in vitro hypersensitivity of progenitors to GM-CSF. A recently activated phase II protocol, coordinated by Drs. Castleberry and Emanuel through the Pediatric Oncology Group, treating newly diagnosed JCML patients with CRA will strive to confirm this drug as a useful new therapy for JCML. Samples obtained through this protocol will aid the laboratory studies aimed at further delineating the links between this clinical CRA response and the in vitro GM-CSF hypersensitivity. The major hypothesis is that CRA, acting through its nuclear receptors, is able to down-regulate the early response genes which are stimulated by the hypersensitive GM-CSF cell response. Further, as an aid to clinicians, this study may prove that GM-CSF hypersensitivity is a useful diagnostic marker for JCML, and, extent of in vitro inhibition of CFU-GM growth by CRA may prove to be a useful predictor of clinical response to CRA. Finally, defining the molecular mechanisms linking the retinoic acid and GM-CSF signaling pathways may lead not only to finding new and more effective therapies for JCML, but also may yield important information concerning the mechanisms underlying or the treatment of other myeloproliferative disorders.